measureResistance Diva

any examples of the use of measureResistance in DIVA ?

Juan F. Osorio
Ph. D. Student
Technical University of Catalonia
http://pmos.upc.es/blues/

 

Juan:
See old post entitled parasitics below:

From: Stefan Bormann ()
Subject: parasitics

View this article only
Newsgroups: comp.cad.cadence
Date: Thu, 15 Mar 2001 14:42:23 +0100

Hi Folks,

I have a problem with parasitic Extraction of resistance and parasitic
capasitance.
I want / must use "measureResistance", but this command allowed the
following parasitic Extraction only with "measureFringe" and
"multiLevelParasitic".
I have diff, poly1, met1, met2, met3. I want measure the capacitance
between this layers and I want measure the resistance of met1-, met2-
and met3-lines.
The problem: after measureResistance disappear layers met1, met2, met3
and I can't measure the parasitic capacitance.

a second problem:
I want measure parasitic resistance selective:
I mark the part of met1 with a label and I aks this label.
now the problem: the shape is true recognize, but it is also disappear
after measure.
no symbol is created..
The false geomConnect via is assign to the nets.
The globel net gnd! (for example) is not exist.

any solutions?

with best regards

Stefan

Message 3 in thread
From: Edward J Kalenda ()
Subject: Re: parasitics

View this article only
Newsgroups: comp.cad.cadence
Date: Thu, 15 Mar 2001 19:56:31 -0800

You are supposed to use the output layer from measureResistance in the
measureFringe and multiLevelParasitic instead of the input layers to
measureResistance. The geomConnect still uses the input layers given to
the measureResistance.

If you use complexParasitic instead of measureFringe and
multiLevelParasitic you can forget about the output layers from
measureResistance and simply use the same layers as input to
measureResistance. The most recent hotfix of 4.4.3 or higher also has
support in complexParasitic for measuring between all layer pairs.

In either case, you have to have recognizable resistor terminals in your
layout. These would be contacts and vias in the geomConnect, device
terminals and top level cell pins.

Message 4 in thread
From: Vandevalk, Gerry [PINE:5T36:EXCH] ()
Subject: Re: parasitics

View this article only
Newsgroups: comp.cad.cadence
Date: 2001-03-19 12:56:04 PST

Yes this is one of the major flaws in measureResistance.

Why?

It is true that the classic measureParasitic now silently fails to work on any
conductor body
that is a parasitic resistor, you can however use one of

multiLevelParasitic() OR complexParasitic ()
with measureFringe()

and get a nominal answer.

I use a bit of trickery to add this capability.

since you have a layer called met1,
I create an auxiliary layer called met1 purpose R.
Now I exclude the met1 R from met1 as it is created.
(this is real tricky becase you must use geomOr( "met1" ) to connect met1
"drawing" to met1 "pin" ,
and then you need to remove met1 "R" from the result.

 

The online manual has basic examples of all the rules. If you need more
than that, you need to let us know what you are trying to do that the
basic examples don't help with.

Your best bet is to use a PDK from one of the foundries or from one of
the universities like NCSU. At least I think it is North Carolina State
University.